During the third year I will continue to characterize the dog's ventilatory response to exercise. I will specifically examine the response to two types of change in the level of exercise. First I will repeat Bainton's observations on the effects of increasing grade to determine whether the hypercapnea (which in my hands accounts for only a portion of the response) is a result of postural adjustments or a true failure of drive. I will concentrate most of my attention on the transition from one speed to another. I find that in this transition greater than 80 percent of the ventilatory response occurs in the first two breaths and if CO2 is prevented from falling the response is maintained. I will quantify both the early and late components of the response and it's accuracy in response to changing levels of grade. I will confirm or deny the observation that this component is insensitive to the level of chemical drive. Finally I will attempt to determine whether the exercise response and particularly the large early component can be altered by appropriate training.